Jones v Michael Vincent Garnett (HM Inspector of Taxes)

This is an income tax appeal by the taxpayer, Mr. Jones, from a decision of the Special Commissioners, reached on the casting vote of the Presiding Commissioner after the two members of the tribunal could not reach agreement. The case arises under a set of provisions, commonly referred to as the settlement provisions, in the Taxes Act 1988 as amended. I will examine the provisions more closely later, but the general thrust of the particular one which is in point in this case is that, if income from property arises under a “settlement”, which term is widely defined to include an “arrangement”, and the “settlor” has in interest in the property, then the income is treated for income tax purposes as the income of the settlor and not as the income of the person whose income it actually is. Further, the settlor is treated for this purpose as having an interest in the property in various situations, one of which is where income from the property may become payable to his or her wife or husband.

In circumstances which I will describe later, each of Mr. and Mrs. Jones owned one share in a company which earned profits by providing Mr. Jones’s personal services in his skilled field to clients. Mr. Jones drew a comparatively small salary from the company, so that it earned profits. The profits were distributed as dividends and Mrs. Jones received half of them. The Revenue’s case is that the structure is an “arrangement” within the settlement provisions; that Mr. Jones was the settlor of it; and that the dividends paid to Mrs. Jones are treated for income tax as Mr. Jones’s income. Further, the Revenue say that the case is not taken out of the settlement provisions by an exclusion for an outright gift by one spouse to the other.

Mr. Jones’s appeal was heard by two Special Commissioner: Dr. Brice presiding, and Miss Powell. Dr. Brice agreed with the Revenue on both issues (first that the settlement provisions applied and, second, that the exclusion for outright gifts by one spouse to the other did not apply). Dr. Brice was therefore in favour of dismissing the appeal. Miss Powell agreed with Mr. Jones on both points and was in favour of allowing the appeal. Regulation 18 of the Special Commissioners (Jurisdiction and Procedure) Regulations 1994 provides that in such a case “the Special Commissioner presiding at the hearing shall be entitled to a second or casting vote”. Dr. Brice exercised the casting vote for the appeal to be dismissed, giving effect to her own view. Mr. Jones now appeals to the High Court.

A procedural question concerning the casting vote was raised before me, but the major questions on the appeal are the substantive questions. On those, for the reasons which I will explain, I agree with Dr. Brice. I therefore consider that the assessments (for three years of assessment) and the amendment made by the Inspector to Mr. Jones’s self-assessment (for one year) were in principle correctly made. Four years of assessment were before the Commissioners: 1996/97 to 1999/2000. For procedural reasons which I need not go into, the assessments for the first three years fell away, so that the Special Commissioners’ decision was only effective for 1999/2000. Only that year is under appeal before me. The case is, I believe, in the nature of a test case and the single year 1999/2000 is adequate for that purpose. For that year of assessment I shall dismiss the appeal.

After Dr. Brice had exercised her casting vote, Mr. Gammie QC, who appeared for Mr. Jones before the Commissioners and before me, wrote to the Commissioners inviting Dr. Brice to reconsider. He urged that she should exercise the casting vote in favour of allowing the appeal, on the theory that, if one Commissioner took the view that a tax charge did not arise, the Presiding Commissioner ought to have recognised that the Revenue had not succeeded in bringing the taxpayer clearly within the charging provisions. Therefore the benefit of the doubt should be given to the taxpayer, notwithstanding that Dr. Brice herself thought that the charging provisions did apply. Dr. Brice declined to change her casting vote.

The procedural issue which thereby arose is not a live one now, because Mr. Gammie does not invite me to remit the case to the Special Commissioners with some sort of direction as to how Dr. Brice should exercise the casting vote. He recognises that the substantive issues are before me and that I am going to decide them. Therefore the outcome of the case is going to turn on my decision on those issues, and whether or not Dr. Brice should have exercised her casting vote as she did is not going to make any difference. Mr. Gammie nevertheless invites me to consider the casting vote issue and to give a ruling about it, by way of guidance for future occasions. I have considered whether to do that, but I have decided not to. As I have said, it is not a live issue on the present appeal to me. In practice it is a situation which is likely to arise only rarely, because practical experience is that a panel of two Commissioners is almost always able to reach agreement. I think that I should leave the question to be determined on a later occasion, if one arises in which it is important. Obviously it could be important in the sort of tax appeal which turns on questions of fact (as to which the decision of the Commissioners is normally conclusive, since an appeal lies only on questions of law). This case is not of that sort. There are no significant points of fact which we are in dispute, and both parties agree that the issue before me is one of law.

I will, therefore, proceed with my judgment on the substantive issues. I begin by summarising the facts, which are substantially undisputed.

Mr. Jones is a specialist in information technology. Until 1992 he worked as an employee of one or more unrelated companies, but his employment came to an end in that year. He and his wife discussed matters, and decided that he would set up his own business. However, on the advice of accountants he established it in corporate form, and by the use of a company jointly owned by himself and his wife. A company called Arctic Systems Limited was acquired from company formation agents. There are only two shares. The Special Commissioners find that Mr. Jones and Mrs. Jones each paid £1 for his or her share. Mr. Jones was the sole director. Mrs. Jones was the company secretary. Mr. Jones did not have a written service agreement with the company, but all his working activities were carried out as a director and employee. The company’s business consisted of providing Mr. Jones’s services to outside users, through agencies, for fees. Mrs. Jones did some work in the company, but all the earning power was his. Mrs. Jones’s role in the business is summarised by the Special Commissioners in paragraph 17 of their decision as follows:

“Mrs. Jones undertook all the book-keeping work, liaised with accountants and the bank, organised business insurance, prepared the value added tax returns and paid the tax; and did the company’s invoicing. She signed off the company’s accounts in her capacity as company secretary. She discussed with the Appellant new contracts and contract renewals and took calls from agencies to arrange appointments for interviews. She sent out the Appellant’s CV as necessary. She worked on average about four or five hours each week on company business.”

I comment that the work which Mrs. Jones did was obviously valuable and the company could not have been operated without it. Nevertheless, there is no doubt that her role was very much subordinate to that of Mr. Jones in any contribution to the profitability of the company.

I should also read two other passages from the decision. Paragraph 14 reads as follows:

“The acquisition by each of the Appellant and Mrs. Jones of one share was recommended by the first accountants who advised them that that was the standard method of working for similar companies. The first accountants also advised that it was normal for husbands and wives to own the shares in this way as the entitlement to dividends depended upon the ownership of the shares. The Appellant understood that, if dividends were paid to his wife, the overall tax payable would be less than it would be if all the dividends were to be paid to him.”

I also read two sentences from paragraph 18.

“As the shareholders of the company the Appellant and Mrs. Jones agreed that the company would pay them salaries which would meet their basic needs and that any profits would be distributed as dividends. Of course, it was the Appellant as director who would decide whether to declare the dividends.”

The company started business in 1992, but the tax issues which are the subject matter of the present appeal were not raised for years of assessment before 1996/97. Further, as I have already explained, the present appeal to this court relates only to 1999/2000. The company’s accounting period ran to 31 October each year, whereas of course the income tax year to which the assessments relate runs to 5 April in each year. Some figures are given in the decision, and the documents include the accounts of the company, some of which are only abbreviated accounts. I have some difficulty in matching figures in the accounts with figures for years of assessment, bearing in mind that the accounting year and the year of assessment do not exactly correspond. Further, even when I would have expected a figure in the decision to be the same as a figure in the accounts, quite often they are not quite the same. Nevertheless, the following summary, which I derive principally from the company’s audited accounts to 31 October 1999 (dividends for which period I would expect to be reflected in the 1999/2000 income tax assessment), should give a broad impression of how matters were conducted. The pattern was, I believe, in essence similar in all other years.

The company’s profit and loss account to 31 October 1999 begins by recording “turnover” of £91,123. I think that this must all have been fees for the provision of Mr. Jones’s services. There are then “administrative expenses” of £16,466. This amount certainly includes the salaries of Mr. Jones and of Mrs. Jones, but there is no breakdown of the £16,466 in the account, so I do not know exactly what their respective salaries for the accounting year were. However, for the tax year 1998/99 (ending on 5 April 1999) the decision records that Mr. Jones’s salary was £6,868 and that Mrs. Jones’s salary was £3,600. The figures for their salaries were quite similar for the following tax year 1999/2000. So I think that in broad terms I can say that something between £10,000 and £11,000 of the administrative expenses, which totalled £16,466, was salaries and the balance must have been other miscellaneous items. The profit and loss account records a net profit before tax (that is after, among things, Mr. and Mrs. Jones’s salaries) of £74,728. £15,265 is provided for corporation tax, leaving a post-tax profit of £59,463. The account then records dividends of £60,535, of which half would be payable to Mr. Jones and half to Mrs. Jones. I mention that the figure recorded in the decision for dividends is not quite the same as the figure in the accounts. I do not know why not, but since I am only concerned with questions of principle in this judgment, I do not need to get to the bottom of that minor puzzle.

The general pattern was, I think, the same for all relevant years, and can be encapsulated as follows. The company received quite large sums for providing Mr. Jones’s services to third parties; it paid small salaries to Mr. and Mrs. Jones, thereby leaving itself with a substantial profit. After providing for corporation tax on the profit, it distributed virtually all the balance as dividends. Reverting briefly to the salaries, the case has proceeded on the footing that the salary to Mrs. Jones (for four or five hours work a week on book-keeping and the like) was adequate. However, the salary paid to Mr. Jones was plainly far less than his expertise was able to generate for the company. The dividends paid were substantially greater than they would have been if Mr. Jones had received a salary commensurate with his earning power. There is of course no requirement of any sort, either in company law or tax law, that Mr. Jones should take a salary commensurate with his earning power. If he chose to draw a low salary, that was entirely proper and open to him. However, his choice might have tax consequences, and in my judgment in this particular case it did. Still on the relationship between salaries and dividends, passages which I have already quoted from the decision show that the pattern which the company adopted derived from advice from accountants, and that Mr Jones understood that the payment of dividends to Mrs. Jones resulted in less tax being paid. (In holding that understanding, Mr. Jones did not of course take into account the possible application of the settlement provisions.) The explanation for it being expected that the tax payable would be lower must have been that Mrs. Jones’s income was such that her dividends (or part of them) would be taxed at lower rates.

It is against that background that the Revenue have contended that the dividends paid to Mrs. Jones were income arising under an “arrangement” of which the settlor was Mr. Jones and that the dividends were deemed to be his income under the settlement provisions.

The settlement provisions have been part of the tax code for many years. They had, however, become rather unwieldy and, in some respects they overlapped with each other. They were significantly recast and tidied up by the Finance Act 1995, but they are still in form sections of the Income and Corporation Taxes Act 1988. The intention of the 1995 Act was to preserve the substance of the old law in most respects. Much of the terminology of the old law is still used, including the terms “settlement” and “arrangement”. Cases decided on the previous provisions are still relevant. I will set out the present provisions in so far as they are relevant to this case. I will then explain how the Revenue say that they apply, and in the course of doing that I will identify which contentions are disputed and which are not. The charging section under which tax is claimed from Mr. Jones on Mrs. Jones’s dividends is section 660A, but I should first refer to critical definitions in section 660G.

Sections 660G(1) and (2) read as follows:

“(1)

In this Chapter -

‘settlement’ includes any disposition, trust, covenant, agreement, arrangement or transfer of assets, and

‘settlor’, in relation to a settlement, means any person by whom the settlement was made.

“(2)

A person shall be deemed for the purposes of this Chapter to have made a settlement if he has made or entered into the settlement directly or indirectly, and, in particular, but without prejudice to the generality of the preceding words, if he has provided or undertaken to provide funds directly or indirectly for the purpose of the settlement ….”

Both of those subsections are identical to earlier provisions which had been in force for many years.

I now move to section 660A. The parts relevant to this case are as follows:

“660A Income arising under settlement where settlor retains an interest.

(1)

Income arising under a settlement during the life of the settlor shall be treated for all purposes of the Income Tax Acts as the income of the settlor and not as the income of any other person unless the income arises from property in which the settlor has no interest.

(2)

Subject to the following provisions of this section, a settlor shall be regarded as having an interest in property if that property or any derived property is, or will or may become, payable to or applicable for the benefit of the settlor or his spouse in any circumstances whatsoever.

….

(6)

The reference in subsection (1) above to a settlement does not include an outright gift by one spouse to the other of property from which income arises, unless -

(a)

the gift does not carry a right to the whole of that income, or

(b)

the property given is wholly or substantially a right to income.

For this purpose a gift is not an outright gift if it is subject to conditions, or if the property given or any derived property is or will or may become, in any circumstances whatsoever, payable to or applicable for the benefit of the donor.

….

(10)

In this section ‘derived property’, in relation to any property, means income from that property or any other property directly or indirectly representing proceeds of, or of income from, that property or income therefrom”.

Under those provisions the arguments are as follows.

It can be seen from the foregoing analysis that there are two questions to be decided. First, apart from the outright gift exclusion in section 660A(6), was there a settlement/arrangement in this case? Second, if there was, is the structure excluded from the description of a settlement/arrangement by the outright gift exclusion?

There have been a number of cases over the years which have explored the ambit of the settlement provisions. Mr. Gammie helpfully and economically took me through fourteen cases, which may be the total of the reported cases. I am not going to refer to all of them, but I will mention those which appear to me to be of specific relevance to the present case. I am particularly concerned with cases which address the meaning of “settlement” and its extension to include an “arrangement”.

The first case was Copeman (H.M. Inspector of Taxes) v. Coleman (1939) 22 TC 594. The specific statutory provision in issue in that case was section 21 of the Finance Act 1936, which defined “settlement” in the same terms as are now found in the present section 660G(1). As Lawrence J said: “The word ‘settlement’ in Section 21 is defined in Sub-section (9), which provides that: ‘the expression “settlement” includes any disposition, trust, covenant, agreement, arrangement or transfer of assets’”. I may be wrong, but I think that the 1936 Act was the originator of the definition. In Copeman v. Coleman, two children of Mr. and Mrs. Coleman each subscribed £10 for a preference share in a company established by Mr. Coleman. They received dividends which were substantially larger than the amounts paid for the shares. In a short judgment Lawrence J held that the structure was a “settlement” as defined, and that Mr. Coleman was a “settlor”. He said this:

“In my opinion, it is impossible to come to any other conclusion but that this was not a bona fide commercial transaction, and it appears to me that there was a disposition within the meaning of the definition in Sub-section (9), or an arrangement in the nature of a disposition within the meaning of that Sub-section. I am also of opinion that the Respondent was a settlor within the meaning of Sub-section (9) (c). I am unable to see how the word ‘indirectly’ can be limited in the way which is suggested so as to exclude the settlements which are made through the interposition of a company”.

The importance of the case for present purposes is that it demonstrates, first, that a “settlement” within the definition can exist without the creation of a trust, and, second, that a corporate structure from which income arises in the form of dividends is capable of being a settlement. As regards the first point, a “settlement” without a statutory definition would in all normal cases require the existence of some form of trust, but the extension of the term by the relevant definition so that it includes a disposition and (especially) an arrangement means that, for the purposes of those provisions, there can be a “settlement” without a trust. As regards the second point, it is of course exceptional for a corporate structure to be a “settlement” as defined, but Copeman v. Coleman showed at an early date that such an analysis was possible.

That last proposition is illustrated and confirmed by two more recent cases which illustrate circumstances in which dividends paid on shares to individual shareholders, without the interposition or existence of a trust, were income arising under a “settlement” as defined. Those cases are Butler (H.M. Inspector of Taxes) v. Wildin (1988) 61 TC 666, and Young (H.M. Inspector of Taxes) v. Pearce (1996) 70 TC 331. Both are decisions of Vinelott J (strictly Sir John Vinelott in Young v. Pearce because the judge heard and decided that case after his retirement as a regular High Court judge). I will not go into the facts, but in each case individuals who were the driving forces behind companies arranged for shares to be issued to member of their families at low subscription prices, and for substantial dividends thereafter to be paid on the shares.

Going back from those two comparatively recent cases to earlier cases, the 1953 decision of the House of Lords in Thomas (H.M. Inspector of Taxes) v. Marshall 34 TC 178, turned on the part of the definition which used the words “transfer of assets”, but it also confirmed that the definition of “settlement” did not require the existence of a trust. The actual decision was that outright gifts by a father to two children of property on which income thereafter arose were “settlements”. Their Lordships rejected an argument that “an absolute and unconditional gift is the antithesis of a settlement and cannot be a ‘transfer of assets’ within the meaning of Section 21”. Lord Morton said:

“The object of the sub-section is, surely, to make it plain that in Section 21 the word ‘settlement’ is to be enlarged to include other transactions which would not be regarded as ‘settlements’ within the meaning which that word ordinarily bears”.

Crossland (H.M. Inspector of Taxes) v. Hawkins (1961) 39 TC 493 is in my opinion important in this case, and I will return to it later. I summarise the case here. One of the settlement provisions in the 1950s provided (as does one of the provisions presently in force) that where, in consequence of a “settlement” as defined, income was paid to an infant child of the “settlor”, the income was treated for tax purposes as the income of the settlor and not of the child. Jack Hawkins was a high-earning film actor of the 1950s. He had three minor children. On the advice of his accountants the following structure was created. Jack Hawkins became an employee of a personal service company at a low salary. The shares in the employer company were owned by trustees of a trust for the benefit of his children. The nominal settlor of the trust was not Jack Hawkins, but his father in law, who settled £100 to create the trust. The employer company contracted to provide Jack Hawkins’s services to a film company. The film company paid a substantial fee to the employer company. The employer company paid a small salary to Mr. Hawkins, leaving it with a large profit on which it no doubt paid tax (in those days not corporation tax, but income tax and profits tax). The employer company then paid one or more dividends to the trustees of the £100 trust which had been established by Mr. Hawkins’s father in law. The trustees paid the dividends on as income to Mr. Hawkins’s minor children.

The Court of Appeal accepted the Revenue’s argument, which Donovan LJ summarised as follows: “… the formation of the company, the service agreement and the deed of settlement together form an arrangement within the terms of [the section which then contained the definition of ‘settlement’], and so are a settlement for the purposes of Section 397 [which corresponds to one of the current sections]. For that settlement, likewise, Mr. Hawkins provided funds and is therefore a settler.” Counsel for Mr. Hawkins had conceded that, if there was such an arrangement, then Jack Hawkins was a settlor of it. Counsel had argued, however, that there was no such “arrangement”. The Court of Appeal held that there was. Donovan LJ said this:

“I think there is sufficient unity about the whole matter to justify it being called an ‘arrangement’ for this purpose, because, as I have said, the ultimate object is to secure for somebody money free from what would otherwise be the burden, or the full burden, of Surtax. Merely because the final step to secure this objective is left unresolved at the outset and decided upon later does not seem to me to rob the scheme of the necessary unity to justify it being called an ‘arrangement’.”

The Revenue say that the present case is effectively covered by the decision in Crossland v. Hawkins. I will not discuss the Revenue’s argument at this point in my judgment, but I say now that I agree with it. There are factual differences between the structure in the present case and the structure in Crossland v. Hawkins, but for reasons which I will explain later, I do not think that they are relevant to the outcome.

Some years after the decision in Crossland v. Hawkins, the principle of that case was affirmed and applied by the House of Lords in Mills v. Commissioners of Inland Revenue (1974) 49 TC 367, [1975] AC 38, another case involving a service company which employed a film actress, and the shares in which were owned by trustees of a trust deed established by another person. The House of Lords held that the whole structure was an “arrangement”, and that the actress was the “settlor” of it.

Finally on the authorities I wish to mention the cases which have established that, before an “arrangement” can be within the definition of “settlement” for purposes of the settlement provisions, an element of “bounty” has to be present. “Arrangement”, after all, is a very wide and general word and could on one view cover a multitude of things which one would not expect to be the subject of provisions deeming the income of one person to be the income of another person for tax purposes. The courts have narrowed the scope of the word “arrangement” in this context by adopting the “bounty” concept. The word “bounty” does not appear in the statute, but it gives a flavour of what the courts are looking for in order to determine whether the settlement provisions apply to a particular structure or not.

I may be wrong, but I think that the first specific use of the word is in the judgment of Plowman J in Commissioners of Inland Revenue v. Leiner (1964) 41 TC 589. I will not attempt to describe the complicated facts of the case and I will simply read one sentence from page 596 of the report.

“The arrangement in my view must be looked at as a whole, and looked at in this way, I find it impossible to say that the Respondent did not provide the trustees with an income of £2,040 a year in the sense in which the word ‘provided’ is used in Section 401 of the Act; that is to say, as importing an element of bounty”.

A couple of years or so later Bulmer v. Commissioners of Inland Revenue (1966) 44 TC 1, [1967] Ch 145, came before Pennycuick J. Members of the Bulmer family, in order to head off an unwelcome takeover campaign directed at their company, made an arrangement (in the general sense of that term) with another company called Sanderson. The arrangement was undoubtedly favourable to Sanderson, and for a time it involved that company receiving dividends on shares in Bulmers. The Revenue argued that the dividends were caught by the settlement provisions. The argument was that the structure was not just an arrangement in the general sense, but also an “arrangement” within the statutory definition, and that the members of the Bulmer family were the settlors of it. Pennycuick J, reversing the Special Commissioners, held that the structure was not an arrangement of the sort to which the definition of settlement applied. The essence of his reasoning was that, so far as the Bulmers were concerned, the structure was put in place for bona fide commercial reasons, and involved no element of “bounty”. I quote just one sentence from page 29 of the report in Tax Cases.

“It seems to me abundantly clear that the transaction between the Appellants and Sanderson was indeed a bona fide commercial transaction. Again, in case that imports in any respect a different test, it is clear that there was no element of bounty as between the appellants and Sanderson”.

The third case, which definitively settles the position that some element of bounty is required for an arrangement to come within the statutory provisions, is Inland Revenue Commissioners v. Plummer [1980] AC 896. The case involved the taxpayer making annuity payments and claiming surtax relief for them. The Revenue argued that the surtax relief was not available because the structure was an “arrangement” within the settlement provisions. The House of Lords held by a majority that the Revenue’s argument failed. The annuity payments were made as part of a tax avoidance scheme. It was true that they were not made for bona fide commercial purposes: they were made for tax avoidance purposes. But equally there was no element of bounty about them. Therefore the settlement provisions did not apply to them. I will read two short extracts from the speech of Lord Wilberforce, one of the majority in the House of Lords.

At this point in my judgment I leave out of account section 660A(6) (the exclusion for an outright gift from one spouse to the other), and I consider whether the dividends paid to Mrs. Jones on her share in Arctic Systems Limited were income arising under an “arrangement” of the sort identified by the authorities. In my judgment, agreeing with Dr. Brice (the presiding Special Commissioner), they were. Mrs. Jones paid £1 for her share. She worked in the business, but did so part time, and she was paid adequate remuneration for the work which she did. On her £1 share she received substantial annual dividends. For the four years of assessments for which the decision gives figures, her annual dividend ranged between a low of £23,450 (in 1997/98) and a high of £28,750 (in 1998/99). All of the receipts of the company, which enabled it to have profits and to pay dividends, were attributable to Mr. Jones, and he drew only very modest salaries. For the four accounting periods ending on 31 October in the years 1997 to 2000 the turnover recorded in the decision ranged from a low of £78,355 (year to 31 October 2000) to a high of £91,123 (year to 31 October 1999). (As I have said earlier, not all of these figures which I have taken from the decision are identical to those in the accounts, but the differences are not great and do not affect the general picture.) Mr. Jones’s salaries for the four years of assessment from 1996/97 to 1999/00 ranged from a low of £6,520 (1999/00) to a high of £7,146 (1999/97). In the circumstances, if there was an arrangement and thus a settlement, there can in my view be no doubt that Mr. Jones was the settlor of it.

In concluding that there was an arrangement, and thus a settlement of which Mr. Jones was the settlor, I am disagreeing with the view of Miss Powell, the second of the two Special Commissioners. Her reasoning was that there was no arrangement because there was no element of bounty at the time when the company was established and when Mrs. Jones acquired her one share in it. At paragraph 96 of the decision Miss Powell wrote this:

“I am persuaded by the Appellant’s argument, presented by Mr. Gammie, that whilst there may have been an arrangement involving that share, not all arrangements are statutory settlements. The arrangement has to be judged at the time the share was acquired by Mrs. Jones and the arrangement at that stage lacked the requisite element of bounty”.

Miss Powell’s point, I think, was that, when the company was established, with Mrs. Jones holding one share and her husband working for the company with a view to earning fees from outside users of his services, her share was probably still worth no more than the £1 which she had paid for it. Mr. Jones was not contractually bound to draw only low salaries from the company, so there could be no certainty that, even if his services did generate large receipts for the company, the result would be significant profits within the company capable of being distributed as dividend. Further, it may have been the case at the original creation of the structure that there was no means of knowing whether the company would succeed in finding outside parties willing to pay substantial fees for Mr. Jones’s services. At paragraph 126 Miss Powell wrote this:

“I conclude that there was no element of bounty in the arrangement when Mrs. Jones acquired her share. The Appellant was not worse off at the time that Mrs. Jones acquired her share merely because he intended to provide bounty. An intention to provide bounty is not the same as the provision of bounty”.

I agree that there is a difference between a present intention to provide bounty and the actual provision of it later. But I do not accept that, if a structure is created with the intention that it shall be a means of providing bounty in future years, it is not an “arrangement” within the meaning of section 660G(1). On the facts Mr. and Mrs. Jones were guided by advice of accountants when they acquired the company. The plan, for all that it may not have been cast in stone and that there could have been changes of mind about it, was that the company would pay to them salaries only at levels which would meet their basic needs, and that any profits would be distributed as dividends. (See the decision, paragraph 18). Mr. Jones understood (and I have no doubt that Mrs. Jones understood as well) that, if dividends were paid to Mrs. Jones (and if the settlement provisions did not apply, no one having contemplated at the time that they might apply), the overall tax payable would be less than it would be if all the dividends were paid to Mr. Jones.

In my view the point of having one share acquired by Mrs. Jones (or at least one of the points) was that she should in future be in a position to receive dividends which, if and when she did receive them, would plainly come to her as bounty. I do not accept Miss Powell’s proposition that, if a structure is established by one person with an intention that bounty will or may flow from it to another person in future, there is not at the outset an arrangement which involves an element of bounty. The word “arrangement” carries to my mind the notion that it comprehends not just the specific things which happen when the arrangement is made, but also the reason or reasons why the arrangement is being made. If a structure is being established in circumstances where one of the reasons for it is that it will or will be available to be used as a means through which bounty will or may be channelled to another person in future, that is in my view fully within what the cases contemplate as an arrangement covered by the statutory definition.

Miss Powell notes that, when Mrs. Jones acquired her share, Mr. Jones was not bound by a service contract to provide his services for only a low level of remuneration. Indeed, Mr. Jones never did become contractually so bound. That is true, but in my judgment it is immaterial in this context. As I have said, parts of the plan or of the intention were that Mr. Jones would draw a low salary and that dividends would be paid (half of them going to Mrs. Jones on her share, which had cost her £1). Therefore in my view those elements of the plan or of the intention were parts of “the arrangement”. They are not prevented from being parts of the arrangement by the feature that they were not legally binding. In many legal documents, whether statutes or private documents, one finds the two words “agreement or arrangement” appearing in conjunction with each other. Indeed, they so appear in the definition of “settlement”, with which I am concerned. (“’Settlement’ includes any disposition, trust, covenant, agreement, arrangement or transfer of assets”). It is, I think, generally understood in instances where the words “agreement or arrangement” are used that “agreement” is likely to mean something which is legally binding, whereas “arrangement” is likely to mean, or at least to include, something which is planned and expected but is not legally binding. This is a further point which, in my judgment, is inconsistent with Miss Powell’s proposition that an intention that a structure created now will be used to provide bounty in future is not enough to make it an “arrangement” within section 660G(1).

Finally on this part of my judgment I wish to return to Crossland v. Hawkins, as I said earlier that I would. I will hypothesise some variants to the actual facts of the case and ask whether they would have made any difference to the result.

First, in the actual case the shares in the company which employed Jack Hawkins were held by trustees of a settlement created by Mr. Hawkins’s father in law. Suppose that they had not been, but that instead the father in law had given £33 outright to each of Mr. Hawkins’s three children in order that the children should be absolute owners of the shares in the company. Nominee holders might have been needed because the children were infants, but there is no doubt that an infant can be the absolute beneficial owner of shares. So on this assumption the dividends would flow directly from the employing company to the children, instead of flowing first to the trustees and then on to the children (as beneficiaries under the trust). Would the result have been any different? My answer is: obviously not.

Second, in the actual case Jack Hawkins had signed a medium term service agreement to be employed by the company at a modest salary. Suppose that it had not been like that, but instead (as is common with many small companies) there was nothing in writing; however, there were expectations that Jack Hawkins would operate as an employee of the company, that he would be paid whatever salary was fixed from time to time, and that in practice salary payments would be set at low levels. Suppose that all of those expectations duly became realities, so that the company in that way made significant profits (doing so because only a modest salary was paid to Jack Hawkins), and that the company paid dividends which went either directly to the three children or to the trustees and then on to the children. Would the tax result have been any different? Again I assert that the answer is: obviously not.

Third, assume that the facts were as in the last two paragraphs, but with the extra feature that shares in the company, instead of being wholly owned by Mr. Hawkins’s three children or by trustees of a settlement for them, were owned as to 50% by Jack Hawkins personally and only as to the other 50% by the three children or by trustees for them. Would the tax result have been the same as respects the 50% of the dividends which flowed to the children, either directly or indirectly via trustees of a settlement created by their maternal grandfather? My answer is: obviously yes. Putting the point the other way round, would the decision in Crossland v. Hawkins have been different if Mr. Hawkins had owned some of the shares in the service company himself. The answer I assert is: obviously not.

Combining the three variants to the facts of Crossland v. Hawkins which I have postulated gives me almost the exact equivalent of this case, with the only difference being that in Crossland v. Hawkins as varied the shares in the employer company not owned by Mr. Hawkins himself were assumed to have been owned by his children, whereas in this case the share in Arctic Systems Limited not owned by Mr. Jones was owned by his wife. In my judgment that is, so far as the present case is concerned, a distinction without a difference. The income of a wife living with her husband is no longer deemed to be her husband’s income for tax purposes: section 279 of the Income and Corporation Taxes Act 1988 was repealed with effect from the year of assessment 1990/91. I interpolate two incidental comments here. First, the previous existence of section 279 meant that the question raised in the present case was rarely of importance. If section 279 was still in force today, Mrs. Jones’s income would be aggregated with her husband’s income anyway by virtue of that section, so there would be no point in debating whether it fell to be so aggregated also by virtue of the settlement provisions. Second, in my view the modern principle that a husband and a wife are taxed separately from each other may mean that it would be more appropriate for the words “or his spouse” not to be included in section 660A(2). After all, if Mr. Jones’s co-shareholder was not his wife, but (say) his sister, he could not be taxed on her dividends. However, the words “or his spouse” are there, and as long as they remain there I agree, so far as this case is concerned, with the decision of Dr. Brice, and with the result which was produced by Dr. Brice’s exercise of the casting vote.

Accordingly, I conclude that, section 660A(6) apart, the Revenue’s claims for tax are correctly made.

I have already set out the subsection, but for convenience I will repeat it here.

“(6)

The reference in subsection (1) above to a settlement does not include an outright gift by one spouse to the other of property from which income arises, unless –

(a)

the gift does not carry a right to the whole of that income, or

(b)

the property given is wholly or substantially a right to income.

For this purpose a gift is not an outright gift if it is subject to conditions, or if the property given or any derived property is or will or may become, in any circumstances whatsoever, payable to or applicable for the benefit of the donor”.

Dr. Brice held that the subsection did not apply, because, in terms of sub-paragraph (b), the property given was wholly or substantially a right to income. Miss Powell, on the other hand, was of the opinion that the subsection did apply and that if (contrary to her view) the case was one of an arrangement of which Mr. Jones was the settlor, the subsection meant that the charging provisions of sections 660A(1) and (2) did not apply.

I agree in the result with Dr. Brice. With some hesitation I agree also with her reason, but in my judgment there is a logically prior point which determines the issue anyway. In my judgment that which constituted the “settlement” here was not an “outright gift” at all. The “settlement” was an arrangement which included the following elements: the acquisition by Mrs. Jones of one share in Arctic Systems Limited for £1; Mr. Jones serving the company as an employee; an expectation that he would draw only a modest salary; and an intention that profits would be paid out as dividends. There was far more comprised in that arrangement than would be covered by the expression “an outright gift”. Indeed, the arrangement did not even include an element which could, even taken in isolation, be regarded as an outright gift. Mr. Jones did not give to Mrs. Jones her share. On the findings of the Special Commissioners she purchased it from the company formation agents, and paid for it with her own money. The money was, of course, only £1, but it remains the case that she got her share because she bought it, not because her husband gave it to her.

Mr. Gammie observes that it could with ease have been done differently: Mr. Jones could have bought both shares in Arctic Systems Limited from the company formation agents, and then given one of them to his wife. I have three comments to make.

Accordingly, I do not consider that section 660A(6) takes the case out of the charging scope of section 660A(1), and I come to that conclusion before I even reach sub-paragraph (b) of subsection (6) on which Dr. Brice bases her conclusion. However, I can see Dr. Brice’s point. Sub-paragraph (b) contains the word “substantially”: “The property given is wholly or substantially a right to income”. What Dr. Brice has in mind, I believe, is that it is unrealistic to view this case as just the gift of one share with a nominal value of £1: it is more complicated than that. I agree, and the conclusion which I would myself draw is that therefore subsection (6) does not begin to apply to this case. However, Dr. Brice has not drawn that conclusion. She has progressed further into the subsection and said that, if one has to look through the share to what was really given, it was the income stream in the form of the annual dividends paid on the shares. I can see the force of that analysis. Personally, I would conclude that the reasons why subsection (6) does not apply are those which I have explained in the previous few paragraphs. But alternatively, or in addition, I would adopt Dr. Brice’s reason, based on the words “substantially a right to income” in sub-paragraph (b) of the subsection.

Whichever reason is the correct one, I hold that section 660A(6) does not take the present case outside the scope of section 660A(1).

For the foregoing reasons, my decision must be that this appeal is dismissed. I would only add this. The decision of the Special Commissioners in this case has attracted considerable attention among professional tax advisers. It seems to have caused much consternation. In my view apprehensions that almost every case of a husband and wife company is going to be affected by this case are greatly exaggerated. If a husband and wife set up a joint company and run it together (for example, the company opens a shop and the couple run and staff it), it does not follow from my judgment in this case that the husband is going to be taxed on the wife’s dividends. It is also an important feature of this case that Mr. Jones provided funds directly or indirectly for the purposes of the settlement (see words appearing in section 660G(2), the definition of “settlor”), by working for Arctic Systems Limited in return for a salary far below his true earning power. It would be far harder for the Revenue to establish that there is a settlement or arrangement of which a husband is a “settlor” if he is paid the going rate for employees carrying out the sort of work which he does.

I have given judgment at some length because there is widespread professional interest in this case, but I do not think that there is anything particularly novel or alarming in my decision. I believe that it is a simple application of well-established principles. Applying those principles, I dismiss the appeal.